摘要
针对目前大规模机群监控系统性能和可用性存在的问题,提出了一种基于IEEE1394协议的机群监控信息采集模型。该模型根据IEEE1394特点,对IEEE1394节点管理模块进行详细分析,实现了用该模块进行监控信息采集,从而替换了Ganglia监控系统中用于监控信息采集的部分。实验结果表明,该模型降低了机群监控系统对机群系统产生的负载,有效地提高了监控信息的采集速率和机群监控系统的可用性。
Aiming at the problems of performance and usability for current large cluster monitor system.A kind of cluster monitoring information-collection model is put forward based on the IEEE1394 protocol.According to IEEE1394 characteristics,this model makes a detailed analysis of IEEE1394 node management module,implements information-collection function and replaces the monitoring information-collection of Ganglia monitor system.The experimental results show that it reduces the load and effectively enhances the monitoring information-collection rate and the usability of cluster monitoring system.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第20期4414-4417,4425,共5页
Computer Engineering and Design
基金
辽宁省博士科研启动基金项目(20051058)